Both P. inflatus (see Johnson and Knight 1966) and P. gracilis (see Báez 1983) have eleven phyllosoma stages which, according to Johnson (1971), are almost identical in almost every respect, except for the presence of a subexopodal spine on the fourth pereopod in stages five through eleven of P. gracilis and the absence of this spine in all stages of P. inflatus. However, molecular analyses have shown that this and other morphological criteria are insufficient to distinguish between phyllosomata of the two species (García-Rodríguez et al. 2008). The length of the larval phase has not been determined for either species.

Johnson (1971) described the puerulus of “P. inflatus-gracilis” from plankton samples collected over an area where the two species co-occur, whereas Báez (1983) described the puerulus of P. gracilis from samples collected in an area where only this species occurs. Both pueruli are similar in size (7.0–8.9 mm carapace length, CL) and have long, spatulated antennae, which are 2.7 times the length of the body in P. gracilis (see Báez 1983) and about 2 times the length of the body in “P. inflatus-gracilis” (see Johnson 1971). Based on these and other minor differences, Báez (1983) suggested that the puerulus of “P. inflatus-gracilis” described by Johnson (1971) belonged to P. inflatus.

Panulirus gracilis occupies different types of benthic habitats, from rocky bottoms with clear water to gravel-sand bottoms near river discharges where water can be considerably turbid, whereas P. inflatus occurs exclusively in rocky habitats with clear waters (Briones et al. 1981, Pérez-González et al. 1992, Pérez-González 2011). Although these findings suggest that P. inflatus might be considered a habitat specialist and P. gracilis a habitat generalist, an important criterion to consider a benthic species as habitat specialist is that its postlarvae settle into the same type of habitat where the adults live, and as habitat generalist that the postlarvae are able to settle in various types of habitats (Adams et al. 2006, Adams and Ebersole 2009); however, the natural settlement habitats have not been determined for either P. inflatus or P. gracilis. However, recently settled pueruli and early benthic juveniles of both species (6-24 mm carapace length, CL) were consistently found co-occurring among the profuse biota fouling the pylons of concrete piers in Zihuatanejo, but those of P. inflatus persist longer in this particular habitat, which resembles the rocky habitats occupied by P. inflatus adults (Briones-Fourzán and Lozano-Álvarez 2013). Also, Gracia and Lozano (1980) found numerous pueruli in the stomachs of benthic catfish Occidentarius platypogon (Günther, 1864) (previously known as Netuma platypogon), which they assigned to “P. inflatus-gracilis” following Johnson (1971). However, it is possible that those pueruli belonged to P. gracilis which, unlike P. inflatus, dwells in the same benthic habitats as O. platypogon (gravel-sand and muddy bottoms).

Adults of P. inflatus and P. gracilis reach a similar body size (Fig. 2A). In Zihuatanejo, P. inflatus has a slightly larger mean size (CL) than P. gracilis (Fig. 2B), but mark-recapture data showed that P. gracilis grows almost twice as fast as P. inflatus (e.g. growth rate for males, P. gracilis: 0.96 ± 0.08 mm CL week^–1, P. inflatus: 0.56 ± 0.10 mm CL week^–1, Briones-Fourzán and Lozano-Álvarez 2003) (Fig. 2C). In this same location, the size of the smallest ovigerous females ever recorded and the size at which 50% of females are ovigerous (CL₅₀) were slightly larger for P. gracilis (47.5 mm CL and 80.0 mm CL, respectively) than for P. inflatus (45.6 mm CL and 77.5 mm CL, respectively) (Weinborn 1977, Gracia 1985, Briones-Fourzán and Lozano-Álvarez 1992). Both species have an extended yearly reproductive period during which individual females can produce up to four clutches (and possible more), with embryo development taking approximately three to four weeks before hatching (Briones-Fourzán and Lozano-Álvarez 1992, Torres-Zepeda et al. 2008). Using original data on clutch size (number of eggs) versus CL concurrently taken for P. inflatus (see Gracia 1985) and P. gracilis (see Fernández-Lomelín 1992) from Zihuatanejo (Fig. 2D), an analysis of covariance (ANCOVA) showed that, after controlling for the significant effect of CL (F_1,82 = 231.71, p < 0.0001), size-specific fecundity is significantly greater in P. gracilis than in P. inflatus (F_1,82 = 16.24, p < 0.001) (Fig. 2D). Large broods are achieved partly through selection for smaller egg size. Pollock (1997) showed that, for spiny lobsters and other crustaceans, the number of eggs per gram of body weight provides an inverse index of egg size (i.e., the larger the index the smaller the egg). The use of this index shows that the eggs of P. gracilis (1047 ± 87 eggs g^–1 body weight) are indeed significantly smaller than those of P. inflatus (911 ± 54 eggs g^–1 body weight) (Student’s t-test, t₈₅ = 2.685, p = 0.009).

Figure 2. 

Differences in some life-history traits between Panulirus gracilis and P. inflatus from Zihuatanejo, Mexico. A carapace length (CL) distribution (n P. gracilis: 2162, n P. inflatus: 1873) B mean size C growth rate of males (mm CL week^–1, n P. gracilis: 148, n P. inflatus: 34) D brood size (number of eggs per clutch) versus CL relationship. Error bars denote 95% confidence intervals. (Data from A, BBriones-Fourzán and Lozano-Álvarez 1992, CBriones-Fourzán and Lozano-Álvarez 2003, DGracia 1985, Fernández-Lomelín 1992).

Stomach content analyses showed that P. gracilis and P. inflatus consume various types of invertebrate prey but that both species exhibit a marked preference for molluscs (Lozano-Álvarez and Aramoni-Serrano 1996) (Fig. 3). In Zihuatanejo, a capture-recapture experiment was conducted during 1979–80 to estimate monthly lobster densities on a 36-ha rocky site (“site A”) where the two species co-occurred (Lozano et al. 1982). At the same time, the seasonal composition of the benthic community at site A and other sites, as well as the seasonal changes in condition factor of the two lobster species were studied (Aramoni-Serrano 1982, Lozano-Álvarez and Aramoni-Serrano 1996). The total density of lobsters on site A showed a marked increase in September-October relative to the other months (Fig. 4A). For each separate species, the density showed values ≤ 15 ind. ha^–1 between April and August, but then more than doubled in September. In October, the density of P. inflatus doubled again while that of P. gracilis decreased to previous levels. By November, the density of P. inflatus also decreased to previous levels (Fig. 4A).

Figure 3. 

Diet of Panulirus gracilis and P. inflatus from Zihuatanejo, Mexico. For each food item the index of relative importance (IRI) is estimated as IRI = (% frequency × % weight)/100. (Data from Lozano-Álvarez and Aramoni-Serrano 1996).

Figure 4. 

Potential ecological interactions between Panulirus gracilis and P. inflatus in a rocky site (“Site A”) in Zihuatanejo, Mexico. A lobster density (number of individuals ha^–1) B relative abundance of molluscs (percentage of molluscs in benthic samples) C condition factor of lobsters. Error bars denote 95% CI. (Data from ALozano et al. 1982, BAramoni-Serrano 1982, CLozano-Álvarez and Aramoni-Serrano 1996).

Interestingly, site A (but not other sites) exhibited a peak in relative abundance of molluscs in the autumn (Fig. 4B), suggesting that this particular site became a food-rich habitat patch for lobsters during this season (Aramoni-Serrano 1982) and potentially explaining the substantial increase in local density of both P. inflatus and P. gracilis in September, which possibly reached the carrying capacity of the site. In October, the increase in density of P. inflatus and decrease in density of P. gracilis was followed by the recapture of several P. gracilis lobsters during October to December on a gravel-sand site 3 km away from site A (Lozano et al. 1982), suggesting that P. inflatus was able to displace P. gracilis from the food-rich patch. However, the high density of P. inflatus during the autumn possibly resulted in intraspecific competition for food resources, further exacerbated by the local decrease in abundance of molluscs during the winter (Lozano-Álvarez and Aramoni-Serrano 1996), as indicated by the significantly lower condition factor of P. inflatus during the winter relative to other seasons and to P. gracilis (Fig. 4C). In conjunction, these findings suggest that P. inflatus is the superior competitor in the rocky habitats to which this species is restricted.

Lewis (1951) described eleven phyllosoma stages for P. argus from plankton samples and estimated the larval duration in about six months, but Goldstein et al. (2008), who obtained the complete larval phase in the laboratory, identified only ten distinct stages with the entire larval duration in these controlled conditions varying between 4.6 and 6.6 months. The early phyllosoma stages of P. guttatus have not been described and the larval duration of this species has not been determined. According to Chitty (1973), first-stage phyllosomata of P. argus and P. guttatus are virtually indistinguishable based on morphology and size. However, stages VI to X of P. guttatus are substantially larger than the corresponding stages of P. argus (Baisre and Alfonso 1994).

The puerulus of P. argus is relatively small (6.1 mm CL on average) and has tapered antennae about 1.5 times the length of the body (Lewis et al. 1952, Goldstein et al. 2008). In contrast, the puerulus of P. guttatus is quite large (10 mm CL) and has long, spatulated antennae about 2.5 times the length of the body (Briones-Fourzán and McWilliam 1997). However, upon molting into the first juvenile stage and as individuals continue to grow, the antennae of P. guttatus become progressively shorter and thinner than those of P. argus (Briones-Fourzán et al. 2006).

It is well known that P. argus is an ontogenetic shifter, i.e. a species wherein the postlarvae settle into habitats distinct from those of the adults and further undergo notable ontogenetic habitat shifts toward the adult habitat (Adams and Ebersole 2009). The pueruli of P. argus settle in vegetated habitats of shallow reef lagoons and bays (seagrass meadows, macroalgal beds, coastal mangroves), where the early benthic juveniles remain for a few months, taking shelter among the vegetation. Eventually, juveniles outgrow the protection afforded by the vegetation and seek shelter in any crevice-type structure in or adjacent to the settlement habitats before gradually moving to the coral reef habitats where the subadults and adults dwell (Butler et al. 2006). Mating and brooding of P. argus occur in the reef habitat, but after embryo development – which takes from three to four weeks – is completed, the females move to deeper areas to release the phyllosoma larvae (Bertelsen 2013), which develop in oceanic waters.

Upon changing habitats, ontogenetic shifters also tend to undergo changes in behavior (Adams et al. 2006). Indeed, after their first benthic habitat shift, P. argus lobsters change from being asocial to being highly gregarious, with multiple individuals commonly sharing individual crevice shelters (Childress and Herrnkind 1996). In addition, P. argus has a highly mobile lifestyle, with movement ranges increasing with lobster size. In some locations, these movements include organized mass migrations over tens to hundreds of kilometers (Herrnkind 1969).

By contrast, P. guttatus is a habitat specialist, as the pueruli of this species settle directly into the coral reef habitat where the juveniles and adults also dwell (Briones-Fourzán and McWilliam 1997). Individuals of P. guttatus are highly sedentary, with a home range for adults of approximately 100 m in radius (Lozano-Álvarez et al. 2002). Therefore, growth, mating, brooding, and egg hatching all take place in the coral reef habitat (Briones-Fourzán and Contreras-Ortiz 1999, Negrete-Soto et al. 2002). Individuals of P. guttatus use reef crevices as shelters, and although small groups can share crevices, many individuals dwell solitarily, reflecting a much lower degree of gregariousness than that exhibited by P. argus (Briones-Fourzán 1995, Sharp et al. 1997, Briones-Fourzán and Lozano-Álvarez 2008).

Adults of P. argus and P. guttatus have a very different body size (Fig. 5A). For example, in the Puerto Morelos coral reef, P. argus has a much larger mean size (82.3 ± 2.24 mm CL) than P. guttatus (59.0 ± 0.83 mm CL) (Fig. 5B) and the former species also grows much faster than the latter (weekly growth rate for males, P. argus: 0.91 ± 0.6 mm CL week^–1, P. guttatus: 0.26 ± 0.13 mm CL week^–1) (Lozano-Álvarez et al. 1991, Negrete-Soto et al. 2002) (Fig. 5C). In the same location, the largest ovigerous female of P. guttatus ever recorded (73.5 mm CL) was smaller than the smallest ovigerous female of P. argus ever recorded (75.0 mm CL). In both species, large females can produce up to four broods per year (Cruz and de León 1991, Briones-Fourzán and Contreras-Ortiz 1999), but the CL₅₀ of ovigerous females is 95.5 mm CL for P. argus and 59.0 mm CL for P. guttatus (Briones-Fourzán 1995). Due to the large interspecific difference in size, size-specific fecundity is far larger in P. argus than in P. guttatus (Fonseca-Larios and Briones-Fourzán 1998, Briones-Fourzán and Contreras-Ortiz 1999) (Fig. 5D), more so when the size of the eggs is taken into account. As Pollock (1997) had previously noted, the number of eggs per gram of body weight is significantly larger (indicating smaller eggs) in P. argus (689 ± 27) than in P. guttatus (519 ± 15) (t₃₂₂ = 10.925, p < 0.0001).

Figure 5. 

Differences in some life-history traits between Panulirus argus and P. guttatus from Puerto Morelos, Mexico. A carapace length (CL) distribution (n P. argus: 717, n P. guttatus: 450) B mean size C growth rate of males (mm CL week^–1, n P. argus: 148, n P. guttatus: 57) D brood size (number of eggs per clutch) versus CL relationship. Error bars denote 95% confidence intervals. (Data from A, BLozano-Álvarez et al. 2007, Briones-Fourzán and Lozano-Álvarez 2013, CNegrete-Soto et al. 2002, DFonseca-Larios and Briones-Fourzán 1998, Briones-Fourzán and Contreras-Ortiz 1999).

The benthic distribution of P. argus and P. guttatus overlaps in the coral reef habitat. In Puerto Morelos, P. guttatus outnumbers P. argus by 5 to 1 across the entire reef habitat, but the relative density of each species varies with reef zone. Thus, the ratio of P. guttatus to P. argus is, on average, 2:1 in the back reef (the protected reef zone facing the mainland), but 16:1 in the fore reef (the exposed reef zone facing the open waters) (Lozano-Álvarez et al. 2007). A numerical dominance of P. guttatus over P. argus on fore reefs has also been reported in Florida (Sharp et al. 1997) and Belize (Acosta and Robertson 2003). However, there is no evidence that P. guttatus can displace P. argus via interference competition because individuals of these congeneric species do not act aggressively toward each other even when in close proximity (Lozano-Álvarez and Briones-Fourzán 2001). Rather, there is evidence that P. guttatus and P. argus make a differential use of reef resources (Lozano-Álvarez et al. 2007). For example, although lobsters of both species feed on a wide variety of organisms with a marked preference for crustaceans and molluscs (Colinas-Sánchez and Briones-Fourzán 1990) (Fig. 6), interspecific competition for food resources is unlikely, as individuals of P. guttatus forage on the reef itself (Wynne and Côté 2007) whereas reef-dwelling individuals of P. argus forage on seagrass and soft-bottom areas adjacent to the coral reefs (Cox et al. 1997, Briones-Fourzán et al. 2003).

Figure 6. 

Diet of Panulirus argus and P. guttatus from Puerto Morelos, Mexico. For each food item the index of relative importance (IRI) is estimated as IRI = (% frequency × % weight)/100. (Data from Colinas-Sánchez and Briones-Fourzán 1990).

Also, P. argus lobsters tend to occupy crevices (‘dens’) along the lower and middle portions of the reef and P. guttatus lobsters over the middle and upper portions of the reef (Lozano-Álvarez et al. 2007). Occupation of individual dens by multiple conspecifics is more common for P. argus, whereas occupation of individual dens by solitary individuals is more common for P. guttatus. Moreover, in dens harboring either species separately or both species together, P. argus lobsters typically occupy the floor and the entrances of the dens, while P. guttatus lobster typically occupy the deep recesses of the dens, clinging to the ceiling or walls (Sharp et al. 1997, Lozano-Álvarez et al. 2007, Briones-Fourzán and Lozano-Álvarez 2008). The differential pattern of occupation of the den space by these congeners appears to minimize competition for shelter and to be related with a differential susceptibility to predators.

An important ecological trade-off that favors local coexistence of similar species within the same trophic level is a differential susceptibility to predators (Amarasekare 2008). This has been explored for P. guttatus and P. argus, which are potential prey to the same predators in the coral reef habitat (e.g. Randall 1967). In field studies, a negative relationship between the abundances of predators and prey would suggest that the abundance of predators controls the abundance of prey (e.g. Sih 1984, Hixon and Beets 1993, Eggleston et al. 1997). In the Puerto Morelos coral reef, Lozano-Álvarez et al. (2007) examined the relationship between the densities of predators and lobsters of each species by reef zone. A negative relationship emerged only for P. guttatus at the fore-reef zone, where this species was numerically dominant.

In the laboratory, individuals of P. guttatus and P. argus differed significantly in performance of several defense mechanisms expressed by spiny lobsters, indicating a differential defense strategy for each species (Briones-Fourzán et al. 2006). In particular, P. argus relies heavily on the formidable defense of its long, strong spiny antennae, especially when multiple individuals join in cooperative defense, whereas P. guttatus, which has much thinner and weaker antennae, does not express communal defensive behavior at all. Panulirus guttatus lobsters are also more cryptic and only emerge from their shelters to forage for short periods during the night. Individuals of P. argus have to traverse open areas to forage and tend to remain stationary to minimize detection when a predator approaches. If a predator attacks, a P. argus lobster can confront the predator, turning deftly to face it at all times, lashing and raking at the predator with its antennae. By contrast, if an individual of P. guttatus detects an approaching predator, it retreats backwards into the nearest available reef crevice, and if attacked, it can swim backwards in a slow but protracted bout that will effectively remove it from the visual field of the predator. In essence, P. guttatus exhibits a shy behavioral type and a higher susceptibility to predators, whereas P. argus exhibits a bold behavioral type and a lower susceptibility to predators (Briones-Fourzán et al. 2006).

The behavior of spiny lobsters is largely mediated by chemical communication (Aggio and Derby 2011). Because shelter is a limiting resource for these lobsters, individuals that are seeking shelter tend to be attracted to chemical scents released by sheltered conspecifics (“attraction cues”). On the other hand, avoiding scents from a lethally injured or freshly killed conspecific (“alarm odors”), which may signal the proximity of a predator, is a particularly effective antipredator strategy for gregarious species (Dicke and Grostal 2001). However, the degree of gregariousness varies widely among spiny lobsters (Childress 2007) and is particularly different between P. argus and P. guttatus (see Briones-Fourzán and Lozano-Álvarez 2008).

Briones-Fourzán et al. (2008) compared how individuals of P. argus and P. guttatus responded to attraction cues and alarm odors from either conspecifics or congeners. As expected, individuals of both species were significantly attracted to shelters emanating conspecific attraction cues but responded neutrally to shelters emanating congeneric attraction cues. However, individuals of P. guttatus responded neutrally to shelters emanating either conspecific or congeneric alarm odors, whereas individuals of P. argus significantly avoided shelters emanating either conspecific or congeneric alarm odors. The differential responses to alarm odors between species suggest that the cryptic defensive behavior of P. guttatus appears to be sufficiently adaptive to offset the need to avoid dens with conspecific (and congeneric) alarm odors, whereas learning to avoid dens with alarm odors from P. guttatus likely increases fitness in reef-dwelling P. argus, which leave their reef shelters to forage elsewhere during the night and then have to return to the reef to shelter during the day (Briones-Fourzán et al. 2008).